ORNL researcher Radha Krishna Moorthy tests Intelligent Power Stage equipment as part of a newly developed power electronics technology suite that can help manage complex electric grid architecture. (Image: ORNL/Carlos Jones, U.S. Department of Energy)
Tech Briefs: Could you tell me something about your system of building blocks?

Dr. Radha Krishna Moorthy: The fundamental building block is what we call SUPER, which stands for smart universal power electronic regulator. You can think of it as either a DC/ DC system — an inverter — or a DC/DC stage coupled to a DC/AC stage. The block can take any form or shape depending upon what sort of power conversion you use. SUPERs in turn have their own fundamental subcomponents. For example, the power stages that make up SUPER are called IPS — intelligent power stages.

Tech Briefs: Can you tell me something about the functions of the SUPER.

Krishna Moorthy: The basic function of SUPER is power conversion say from AC to DC or DC to DC. SUPER does it more proactively and hosts a variety of special features from embedded intelligence and decision making, to online health monitoring, autonomous operation capability, and cyber physical security.

If you think about a regular power electronic converter, and if you read reports from solar panel installers or utility personnel, the biggest potential failure points are capacitors and the power semiconductor devices. So, it would be really helpful if you could add an additional layer of monitoring to get more data on the capacitors or to get more data on the semiconductor devices. It can help you analyze the health of the system.

Tech Briefs: What kinds of measurements would you take to measure the health of the semiconductors?

Krishna Moorthy: We measure the on-state voltage of the semiconductor device. And we have provisions to directly measure the drain source. So, using these two, I can compute the on-state resistance, which could help me predict the failure mechanism related to semiconductor modules. And besides that, there is data like base plate temperature, junction temperature of the devices, capacitance estimates, etc. We can accumulate this data at an upper layer to give us information about the system health.

Tech Briefs: And how about the health of the capacitors?

Krishna Moorthy: One way to judge capacitor health is to measure the ripple current to see how it is changing. The other way is to continuously compute the capacitance in the system. For example, if the system goes into standby, you can do a simple RC time constant-based calculation to determine the system capacitance over a course of time. So, if anything has failed or any capacity has degraded, or it’s moving toward degradation, you would get to know in advance.

Tech Briefs: What sort of power ratings are we talking about?

Krishna Moorthy: In the ones we demonstrated, we were talking about a 75 KVA system at 480 volts AC, but these can be scaled for higher power and higher voltages.

Tech Briefs: Can you tell me a bit more about the overall architecture.

Krishna Moorthy: The basic inverter is the SUPER. It can be made up of its own fundamental power stages, with more subcomponents on top of it. The IPS is one of the subcomponents of the super. It's basically a holistic power stage with input, output, side contactors, the semiconductor devices, the associated gate drivers, the smaller passives, the associated passives that they need, and the advanced sensing entities within it. So, IPSs can do certain control functions, but can be coordinated to act as one inverter from the top using the SUPER framework.

To put it in the simplest terms: If, for example, I want to build a 250-kilowatt inverter, it will have, say, five IPSs that each are rated for 50 kilowatts, as well as other subcomponents including passive elements, auxiliary power circuits, etc. These in turn will work together to deliver the 250 kilowatts.

Our guiding philosophy is that if one IPS fails, I will still be able to operate the others rather than bringing the whole system down, although the total output power will be reduced.

Tech Briefs: So, does the IPS include all the regulation and controls?

Krishna Moorthy: The IPSs only do the pulse width modulation (PWM) generation The SUPER coordinates them and regulates their outputs and inputs and provides other necessary functions.

Tech Briefs: It said in the press release that a problem with traditional systems is you can typically only repair them with parts from the original manufacturer, but in your system, that's different.

Krishna Moorthy: One of the aspects we were looking into was standardization of mechanisms and subcomponents. When you do that, it opens opportunities for vendors to provide subcomponents, rather than building the whole converter. For example, if I have a 250-kilowatt SUPER, I can have another vendor make the IPS.

This is also the construct we are operating on here at the labs. Someone else can do the IPS for me as long as it's standardized so it can coordinate with the SUPER. I can get IPSs and subcomponents as well, from different vendors and plug them together. You can use one vendor for the inverter and others for the internal components.

Tech Briefs: I see. So, you're basically generating a spec that people work to.

Krishna Moorthy: Yes.

Tech Briefs: Could you give me a hypothetical application for this.

Krishna Moorthy: The application I was thinking of is photovoltaic (PV) systems.

If you want, say, a one-MW utility-scale PV, if one of the PV submodules fails, that brings the whole plant down. So rather, if I have an architecture to work around it and that can tell me that this PV is degrading or there's some component failure that's going to occur, I can schedule maintenance around that, rather than bringing the whole system down.

Tech Briefs: How do you maintain a section without shutting down the whole system?

Krishna Moorthy: That's where the architecture comes in. We have layers and layers of hierarchy. So as the IPSs are coordinated by the SUPERS, the SUPERS are coordinated as nodes. We are enabling intelligence at each layer to tell us if something is failing. And then we have layers at the top level to say, “OK, if this is failing and I can operate smoothly without having this, I can derate the output power and work around it.” That is the sort of intelligence we are embedding in every layer.

Tech Briefs: So, are you are you visualizing this at substations or micro grids or something else?

Krishna Moorthy: Honestly, I'm visualizing this everywhere.

Tech Briefs: Do you envision this to be used like microgrids, where you have a single point of connection to a localized sub grid and then you can have different loads and different sources inside the microgrid.

Krishna Moorthy: A microgrid aggregates a number of inverters or converters together, for example, if it's an AC or DC microgrid. SUPER can be used for microgrids, but there's one more level of aggregation you can do — you can aggregate a number of power stages to increase the output power and still have one point of connection to the grid. That's the level of aggregation this architecture is capable of.

Tech Briefs: What do you see as the next step in this project? Where's it going? When am I going to start seeing these out there?

Krishna Moorthy: You will start seeing some of the bigger integration things out there sooner. With regards to this project, we preliminarily define the construct of what these SUPERs can do and the different ways the architecture is helpful.

So now, the bigger thing that I am working on is to integrate them in a node like a microgrid and see how beneficial they are to the micro grid at the top level. Also, to see the sorts of control flexibilities I can get with the greater amount of data that I'm getting downstream.